DISCUSSION OF RESULTS. 213 



carry with them as they move through the solution than they have in the 

 aqueous solutions, their velocity becomes much greater. These two factors 

 both act in the same direction, i. e., they both tend to increase the ionic 

 velocity, and are quite sufficient to justify the conclusion that the molecular 

 conductivities of potassium sulphocyanate are greater in acetone than they 

 are in water, because of a very great increase in the velocity of the ions in 

 the acetone solutions. 



NEGATIVE VISCOSITY COEFFICIENTS. 



It has been known for a number of years that when certain salts are dis- 

 solved in water the resulting solutions have a smaller viscosity than the pure 

 water. Several theories have been proposed to explain this phenomenon. A 

 brief but comprehensive outline of these theories has been given by Jones 

 and Bingham, 1 and a mere reference to the literature will suffice in this 

 connection. Euler 2 employed the " electrostriction theory " of Drude and 

 Nernst 3 as a probable explanation of the negative viscosities ; but it was 

 later shown by Wagner 4 that this theory is incorrect, because the vis- 

 cosity of a solvent may be lowered by the addition of certain non-electro- 

 lytes. Dunstan, 5 Blanchard, 6 Varenne and Godefroy, 7 Thorpe and Rodger, 

 and Traube 8 all seem to attribute the abnormalities in viscosity to the pres- 

 ence of hydrates. The hydrate work of Jones and his co-workers has clearly 

 shown that potassium chloride and similar salts are but little hydrated, even 

 in dilute solutions; and since potassium chloride is one of the salts which 

 produces a marked negative viscosity when dissolved in water, it is very diffi- 

 cult to see how hydrates could enter into the question of negative viscosity 

 to any appreciable extent. Wagner 9 has made a study of the effect on the 

 viscosity of water of about forty-five inorganic salts. From his data we find 

 that the only salts which produce negative viscosity are the salts of caesium, 

 rubidium, potassium, and thallium (in the thallous conditions), viz, caesium 

 chloride, rubidium chloride, potassium chloride, potassium nitrate, and thallous 

 nitrate. All potassium salts do not have this property of diminishing the vis- 

 cosity of water. Potassium sulphate, potassium ferrocyanide, potassium fer- 

 ri cyanide, and potassium chromate, all give positive viscosity coefficients. But 

 this is not at all surprising, since, as has already been mentioned, it was clearly 

 shown that the viscosity of a salt solution is an additive function of the metallic 

 and the non-metallic ions of the dissolved salt. In other words, the cations and 

 anions seem to work counter to each other in some cases, such as potassium sul- 



Loc. cit. 7 Compt. rend., 137, 992 (1903); 138, 



'Ztschr. phys. Chem., 25, 536 (1898). 990 (1904). 



3 Ibid., 15, 79 (1894). 8 Phil. Trans., 185A, 307 (1894). 



4 Ibid., 46, 867 (1903). 9 Ztschr. phys. Chem., 5, 31 (1890). 

 8 Ibid., 49, 590 (1904). 



6 Journ. Amer. Chem. Soc., 26, 1315 (1904). 



